linux_dsm_epyc7002/arch
Radim Krčmář 7099e2e1f4 KVM: VMX: disable PEBS before a guest entry
Linux guests on Haswell (and also SandyBridge and Broadwell, at least)
would crash if you decided to run a host command that uses PEBS, like
  perf record -e 'cpu/mem-stores/pp' -a

This happens because KVM is using VMX MSR switching to disable PEBS, but
SDM [2015-12] 18.4.4.4 Re-configuring PEBS Facilities explains why it
isn't safe:
  When software needs to reconfigure PEBS facilities, it should allow a
  quiescent period between stopping the prior event counting and setting
  up a new PEBS event. The quiescent period is to allow any latent
  residual PEBS records to complete its capture at their previously
  specified buffer address (provided by IA32_DS_AREA).

There might not be a quiescent period after the MSR switch, so a CPU
ends up using host's MSR_IA32_DS_AREA to access an area in guest's
memory.  (Or MSR switching is just buggy on some models.)

The guest can learn something about the host this way:
If the guest doesn't map address pointed by MSR_IA32_DS_AREA, it results
in #PF where we leak host's MSR_IA32_DS_AREA through CR2.

After that, a malicious guest can map and configure memory where
MSR_IA32_DS_AREA is pointing and can therefore get an output from
host's tracing.

This is not a critical leak as the host must initiate with PEBS tracing
and I have not been able to get a record from more than one instruction
before vmentry in vmx_vcpu_run() (that place has most registers already
overwritten with guest's).

We could disable PEBS just few instructions before vmentry, but
disabling it earlier shouldn't affect host tracing too much.
We also don't need to switch MSR_IA32_PEBS_ENABLE on VMENTRY, but that
optimization isn't worth its code, IMO.

(If you are implementing PEBS for guests, be sure to handle the case
 where both host and guest enable PEBS, because this patch doesn't.)

Fixes: 26a4f3c08d ("perf/x86: disable PEBS on a guest entry.")
Cc: <stable@vger.kernel.org>
Reported-by: Jiří Olša <jolsa@redhat.com>
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
2016-03-08 12:46:46 +01:00
..
alpha dma-mapping: always provide the dma_map_ops based implementation 2016-01-20 17:09:18 -08:00
arc arc: SMP: CONFIG_ARC_IPI_DBG cleanup 2016-02-24 14:15:39 +05:30
arm ARM: SoC fixes 2016-03-06 14:14:54 -08:00
arm64 arm64 fix: 2016-03-04 17:43:40 -08:00
avr32 dma-mapping: always provide the dma_map_ops based implementation 2016-01-20 17:09:18 -08:00
blackfin dma-mapping: always provide the dma_map_ops based implementation 2016-01-20 17:09:18 -08:00
c6x dma-mapping: always provide the dma_map_ops based implementation 2016-01-20 17:09:18 -08:00
cris Merge branch 'akpm' (patches from Andrew) 2016-01-21 12:32:08 -08:00
frv dma-mapping: always provide the dma_map_ops based implementation 2016-01-20 17:09:18 -08:00
h8300 Merge branch 'akpm' (patches from Andrew) 2016-01-21 12:32:08 -08:00
hexagon dma-mapping: always provide the dma_map_ops based implementation 2016-01-20 17:09:18 -08:00
ia64 [IA64] Enable copy_file_range syscall for ia64 2016-01-22 14:20:01 -08:00
m32r m32r: fix build failure due to SMP and MMU 2016-02-05 18:10:40 -08:00
m68k m68k/defconfig: Update defconfigs for v4.5-rc1 2016-02-01 10:35:38 +01:00
metag dma-mapping: always provide the dma_map_ops based implementation 2016-01-20 17:09:18 -08:00
microblaze dma-mapping: always provide the dma_map_ops based implementation 2016-01-20 17:09:18 -08:00
mips MIPS: traps: Fix SIGFPE information leak from do_ov' and do_trap_or_bp' 2016-03-04 22:52:32 +01:00
mn10300 dma-mapping: always provide the dma_map_ops based implementation 2016-01-20 17:09:18 -08:00
nios2 dma-mapping: always provide the dma_map_ops based implementation 2016-01-20 17:09:18 -08:00
openrisc dma-mapping: always provide the dma_map_ops based implementation 2016-01-20 17:09:18 -08:00
parisc parisc: Wire up copy_file_range syscall 2016-03-01 23:21:11 +01:00
powerpc powerpc fixes for 4.5 #4 2016-03-06 11:08:06 -08:00
s390 s390/fpu: signals vs. floating point control register 2016-02-22 09:29:35 +01:00
score
sh sh: fix smp_store_mb for !SMP 2016-01-26 10:18:29 +02:00
sparc Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/sparc 2016-03-01 08:57:34 -08:00
tile dma-mapping: always provide the dma_map_ops based implementation 2016-01-20 17:09:18 -08:00
um um: Export pm_power_off 2016-03-05 22:17:52 +01:00
unicore32 dma-mapping: always provide the dma_map_ops based implementation 2016-01-20 17:09:18 -08:00
x86 KVM: VMX: disable PEBS before a guest entry 2016-03-08 12:46:46 +01:00
xtensa dma-mapping: remove <asm-generic/dma-coherent.h> 2016-01-20 17:09:18 -08:00
.gitignore
Kconfig dma-mapping: always provide the dma_map_ops based implementation 2016-01-20 17:09:18 -08:00